Translation, Adaptation, and Cross Language Validation of Short Schema Mode Inventory (SMI)

Received 01 Mar, 2012

Accepted 18 Feb, 2013

Published 30 Jun, 2013

Schema Focused Therapy (SFT) is extensively used in many clinical and forensic institutes. Schema modes represent momentto- moment emotional states and coping responses experienced by everyone. Schema Mode Inventory was developed for mode assessment. This study aims to translate and adapt English version of Schema Mode Inventory (SMI) into Urdu language and to assess its reliability and validity for people in Khyber Pukhtoon Khwa. Oblique translation technique was used for this purpose. Urdu version and the original SMI English version were administered on bilingual subjects (N = 200) including an equal number of men (n = 100) with mean age years (M = 23; SD = 4.4) and women (n = 100) with mean age years (M = 22 years; SD = 1.5), all being post graduate students of Peshawar University (Pakistan). Reliability of both versions was determined by computing Cronbach alpha coefficients and test retest techniques. Item-sum correlations for each subscale as well as inter scale correlations were obtained to establish validity of SMI Short Form. Cross language validation of SMI was also established on an independent sample comprising ninety seven participants. Findings suggest that SMI can be used as a valid and reliable measure for assessment of schema modes.

According to Young, Klosko, and Weishaar (2003) Schema therapy is a form of Cognitive Behavior Therapy (CBT) and especially recommended for treatment of personality disorders. Schema Focused Therapy is an integrative systematic approach for a range of chronic, difficult, and characterological problems (Berstein, Arntz, & de Vos, 2007). This approach involves four basic concepts i.e., early maladaptive schemas, schema domains, coping styles, and schema modes. However, the focus of schema focused therapy is on lifelong, self-defeating prototypes called early maladaptive schemas (Young & Klosko, 1994).

Schema modes (a set of schemas) describe an individual’s current predominant emotional state and cognitive and coping responses, often triggered by life situations or scenarios to which people are oversensitive (Bamber, 2008). In other words, schema modes become activated when particular schemas or coping responses develop into strong emotions that significantly influence an individual’s functioning (Perocelli, Glaser, Calhoun, & Campbell, 2001). These rapid variations in emotional and behavioral states of patients suffering from severe personality disorders are called schema modes (Lobbestael, Van Vreeswijk, & Artnz, 2007). These Schema modes develop early in childhood and are shaped by the selective filtration of incoming experiences and continue to mature and change throughout the life span (Lobbestael, Arntz, & Sieswerda, 2005).

Young et al. (2003) identified ten schema modes and grouped them into four general categories i.e., maladaptive child modes, dysfunctional coping modes, dysfunctional parent modes, and healthy adult modes. Maladaptive child modes develop when certain basic emotional needs are not adequately satisfied during childhood. As a result, the dysfunctional coping modes appear to reflect an overuse of the coping styles of overcompensation, avoidance, or surrender. For instance, maladaptive parents’ modes reflect internalized behavior of the parents towards the individual as a child while a healthy mode includes functional cognitions, thoughts, and behaviors, and depict an adaptive coping style.

Most individuals use different schema modes in response to a situation or while interacting with people. However, individuals having long lasting characterological problems utilize some of the schema modes that are maladaptive, for instance angry child, impulsive/undisciplined child, compliant surrender, detached protector, over-compensator, punitive parent, and demanding parent (Lobbestael et al., 2007). Young et al. (2005) conceptualized a theoretical link, which indicated that individuals frequently change schema modes in response to varying situational changes. Consequently, most of the healthy individuals not only exhibit several different schema modes, they often change them. However, individuals with a range of psychopathology including Axis-I and Axis-II psychiatric disorders are often dominated by excessive utilization of a single maladaptive schema mode (Lobbestael et al., 2005; Young et al., 2005) and they often stick rigidly to these maladaptive schema modes (Berstein et al., 2007; Young, 1990) as they experience difficulty in utilizing different schema modes simultaneously (Bamber, 2008).

Consequently, they may engage in maladaptive and rigid coping behaviors (Arntz, Klokman, & Sieswerda, 2005). Young et al. (2007) devised Schema Mode Inventory (SMI) comprising 270 items, a comprehensive self-report scale for the assessment of 16 schema modes. Later on Lobbestael et al. (2007) developed a short version of SMI comprising 118 items, which is less time consuming and can be readily used for individuals with different cultural backgrounds. The authors established the reliability and validity of this short form by administering this scale on diagnosed patients on Axis 1 and Axis II disorders of DSM IV-TR and non patients.

Empirical evidence shows that short version is a psychometrically sound measure that can be used for mode assessment in schema focused therapy (Lobbestael, Van Vreeswijk, Spinhoven, Schouten, & Arntz, 2010). This inventory is available in English and Dutch language. In order to use SMI for schema mode assessment in Pakistan, the researchers translated this inventory in national language i.e., Urdu. The addition of this inventory into the practice of clinical psychologists in Pakistan will provide an opportunity for a profound understanding of maladaptive schema modes of individuals having characterological problems. The early identification of maladaptive modes would enable the clinicians to address them through schema focused therapy intervention in time. The investigation of maladaptive schema modes would also help the clinicians working with other psychiatric disorders in order to distinct the dysfunctional modes from the psychiatric symptoms. This will be a step forward in the course of psychological intervention.

METHOD

Schema Mode Inventory (SMI)
Short version of SMI contains 124 items scored on a six-point Likert scale ranging from Never or almost never = 1 to Always = 6. On scales number 1(Vulnerable Child); 2(Angry Child); 3(Enraged child); 6(Happy Child); 10(Self-aggrandizer Child); 12(Punishing Parent); 13(Demanding Parent); 14(Healthy Adult) scores range from 10-60. Each one of these scales comprises ten items. On scales numbers 4(Impulsive Child); 8(Detached Protector); 11(Bully and Attack Mode Child) scores range from 9 - 54. Each one of these three scales consists of 9 items. Scale number 7(Compliant Surrender) consists of 7 items and the score range is 7-42. Scale number 5 (Undisciplined child) consists of 6 items with a score range 6-30. Whereas, scale number 9(Detached Self-Soother Child) consists of 4 items and the scores ranges from 4-24. This short version comprises 4 domains i.e., Child, Maladaptive Coping, Maladaptive Parent, and Healthy Adult and includes 14 schema modes (see Table 1). There are no cut off scores for different subscales/modes and the scores are interpreted by comparing mean score for each mode for the client with the original scoring key. SMI has good discriminant validity and moderate convergent validity, and has good internal consistency for all sub scales α values range between .76 - .96, with a mean value of .86 (Lobbestael et al., 2010).

Translation of Schema Mode Inventory (SMI) Short Form
Before the commencement of the translation process, the authors (Lobbest et al., 2007) were requested to grant permission of translation of Schema Mode Inventory (SMI) Short Form in Urdu Language. The authors graciously accorded the permission of Urdu Translation of this scale. Due to specific sentence structure and constructs used to represent certain personality characteristics and attitudes of the respondents, a word-for-word translation of SMI from English into Urdu language was not possible. It is a common practice to use Oblique Translation Techniques (Mason, 1994), when the structural or conceptual elements of the source language cannot be directly translated without altering meaning or upsetting the grammatical and stylistic elements of the target language. In the present study, the following oblique translation techniques were used, which require in-depth knowledge of source language as well as target language for the translator. Four bilingual experts and the researcher worked together in the Urdu translation of SMI.

Transposition. This method was used to replace certain words in the target language (Urdu) without altering the meaning of the source text (English) for instance items no. 1, 6, 19, 24, 33, 34, 36, 38, 39, and 43.

Reformulation of equivalence/adaptation. Certain items were slightly reformulated to adapt them to respondents’ self-perception/evaluation. The translation of SMI was carried out individually by four experts, who are faculty members of department of English of a public sector university in Khyber Pukhtun Khawa, Peshawar. After completion of the translation process, the draft Urdu versions of SMI were placed before a committee including four bilingual experts and a senior professor in psychometrics. This committee of experts, after a detailed discussion about the appropriateness and equivalence of each item in original SMI and its translation, prepared the Urdu version of Schema Mode Inventory.

Sample
To determine the psychometric properties of SMI (Urdu and English versions) for the present sample, it was administered to 300 participants divided into three groups hereafter called Group I, II & III. The participants of group I and II were students of Peshawar University, Khyber Pukhtoon Khwa, Pakistan and represented normal/non-clinical population. This sample consisted of all the students present on the day of testing and who volunteered to take this test. Total number of respondents was 200 including an equal number of boys and girls. To determine the reliability of SMI and its subscales, the respondents were randomly assigned to two groups, hereafter called Group I and Group II. Each group consisted of 50 boys and 50 girls. However, while scoring the responses it was found that 5 girls and 4 boys comprising group I, did not answer all the questions. Nine answer books were discarded and the remaining 91 respondents including boys (n = 46) and girls (n = 45) comprised groups I. Similarly, 11 respondents who took SMI English version left many items unanswered. As a result the final analysis of responses given by group II was based on 89 cases including boys (n = 44) and girls (n = 45). Both the groups were matched in terms of age, education and socio economic status. Mean age of participants comprising Group I was 22 years (SD = 1.5) whereas mean age of students included in Group II was 23 years (SD = 4.4). All the respondents were from middle socio economic status and were students of Law (LLB). To determine the cross language validity of SMI (English & Urdu versions) another group (Group III) comprising of 100 women students of Shaheed Benazir Bhutto Women University Peshawar, was randomly selected by lottery method from four Post Graduate Departments, namely, statistics, psychology, education, and MBA.

Procedure
The test (SMI) was administered in a group setting. The examiner introduced herself to the group and explained the purpose of the test. After getting verbal consent of the participants, Test booklets were distributed amongst the participants. They were asked to provide personal information by filling in the blanks provided at the top of first page of the test booklet. The examiner explained the procedure to be followed for recording the responses on scale. Urdu version of SMI was administered on Group 1, English version of SMI was administered on Group II and both versions of SMI were administered on group III (N = 100). During the scoring process the researchers found that three respondents comprising Group III did not complete the test and hence their record was not included in the final analysis.

RESULTS

Reliability of SMI
Reliability of both versions (English and Urdu) of SMI was determined by internal consistency and test-retest techniques. To compute alpha coefficients showing internal consistency of SMI Urdu version, it was administered on group I, whereas the original scale was administered on group II. The alpha reliability coefficients are presented in tables 2-3. Furthermore, test-retest reliability (with an interval of one week) of Urdu SMI was found to be .63 (p < .002) whereas for English version it was .87 (p < .000).

Cross Language Validity of SMI
Both versions of SMI (English version and its Urdu translation), were administered with an interval of one week, on Group III comprising 100 women students of Frontier Women University Peshawar, according to the following experimental designs.

Table 1:	Cross Language and Test-retest Reliability of Schema Mode Inventory
**p < .01

Table 1 shows correlation between SMI Urdu and SMI English versions are significant (p < .01). The highest correlation value exists on English version and the second highest correlation value exists on Urdu version.

Table 2:	Cronbach Alpha Reliability Coefficients of Schema Mode Inventory (SMI) Urdu Version
Note. SMI = Schema Mode Inventory

The internal consistency of the SMI Urdu version is acceptable for 9 subscales and alpha coefficients range from .63 to .82 with a mean of .74. However, on 5 subscales viz., Undisciplined Child; Compliant Surrender; Detached Self-Soother; Self aggrandizer; Bully and Attack, the reliability falls below the acceptable index (ranging from α = .41 to α = .56, with a mean of .45). Nevertheless, the overall reliability of SMI Short From is high (α = .92).

Table 3:	Cronbach Alpha Reliability Coefficients of Schema Mode Inventory (SMI) English Version
Note. SMI = Schema Mode Inventory

Table 3 shows reliability of SMI English Version (Short Form) computed for the sample of the present study. The data show that Cronbach alpha of all scales except two (Undisciplined Child and Compliant Surrender) are moderate. Furthermore, correlations of individual items with total scores on the test are significant.

Table 4:	Range of Item-total Correlations on Subscales of SMI

To establish construct validity of SMI (Urdu version) correlation of each item with the total score on each scale (mode), was computed. The range of these correlations is presented in table 4. All correlations presented in table 4 are significant (p < .01). These findings suggest that each item in every scale (other than item no. 31 from Self-aggrandizer child mode and item no. 20 in Healthy Adult mode) is significant suggesting that all the scales are valid measures comprising Urdu version of SMI. On the other hand, item no. 7 in Demanding Parent mode and item no. 73 Healthy Adult mode (English version) do not seem to be contributing meaningfully to the constructs. By and large the original SMI as well as its Urdu translation seem comprising subscales which are valid measures of the constructs involved. Table 5

Table 5:	Inter-scales Correlations of SMI Urdu Version
Note. SMI Scales: 1 = Vulnerable Child; 2 = Angry Child; 3 = Enraged Child; 4 = Impulsive Child; 5 = Undisciplined Child; 6 = Happy Child; 7 = Compliant Surrender; 8 = Detached Protector; 9 = Detached Self-Soother; 10 = Self Aggrandizer; 11 = Bully and Attack; 12 = Punitive Parent; 13 = Demanding Parent; 14 = Healthy Adult. *p < .05. ** p < .01

Table 5 presents inter-scales correlations of SMI Urdu Version. The data show high correlations among eight subscales, i.e., Vulnerable Child, Angry child, Enraged Child, Impulsive Child, Undisciplined Child, Detached Protector, Bully and Attack Mode Child, Punishing Parent. The data further reveals that Happy Child is significantly correlated with Compliant Surrender, Detached Self-Soother, Self Aggrandizer, Bully and Attack, Demanding Parent and Healthy Adult, but an inverse correlation with Punitive Parent. Compliant Surrender shows significant correlation with Detached Self-Soother, Self Aggrandizer, Demanding Parent and Healthy Adult. Similarly Demanding Parent and Healthy Adult are significantly correlated with each other. The correlation matrix for SMI English version demonstrates almost similar pattern.

Comparison of Schema Modes of Group 1 and Group 2
Paired Samples t-test was used to determine if any significant difference exists between the schema modes of the two groups (Group I and Group II) when compared on the basis of their scores on Urdu and English versions of SMI Short Form (t = .796; df = 88). These findings reveal nonsignificant difference between scores of the two groups suggesting the equivalence between the two versions.

DISCUSSION

The Schema Mode Inventory (SMI) was published in 2007 for the first time. Its translation in other languages is continuing, but its reliability and validity has been examined in very few studies (Lobbestael et al., 2010). The present study was carried out in an effort to translate SMI into Urdu language, which is one of the major languages of subcontinent and is national language of Pakistan, the 9^th populous country in the world. Oblique Translation Techniques (Mason, 1994) were used for this purpose. This Inventory was administered on subjects having bilingual background. Our results indicated good reliability and construct validity for both English and Urdu version. These findings suggest that SMI Urdu version seems suitable for the assessment of schema modes in Pakistan. However, during the translation process some statements were found incompatible with the Pakistani culture and society. For this purpose we used the method of transposition. During translation of the inventory, certain words in the target language were replaced without changing the meaning of the source text, and subsequently reformulated slightly and made compatible with the original text (Mason, 1994). We also found the reliability of few individual scales such as Undisciplined child, Complaint Surrender, Detached Self-Soother, Self-Aggrandizer and Bully and Attack modes below the acceptable range, which could be attributed to the differences in child rearing practices, culturally unique role of authority figure, conceptual difference in the context of protection and extreme loyalty to parents and culture, and a cultural difference in the bully and attack happenings. However, the overall results indicated that the Urdu Version of SMI has good reliability and construct validity and could be satisfactorily recommended for the assessment of schema modes. Subsequently, the SMI can be viewed as a viable alternative to the other psychological assessment tools for understanding an individual’s psychological problems. The SMI has the advantage of clearly demonstrating individuals’ maladaptive coping styles, e.g., compliance, surrender, dependence, disconnection, emotional withdrawal, counterattack, self-sacrifice, punitive behaviour, and subjugation, all of which can then be discussed and perhaps targeted for treatment or support by the clinician (Lobbestael et al., 2007). Schema Focused Therapy (SFT) has demonstrated its effectiveness for the treatment of a range of psychiatric disorders particularly Borderline Personality Disorder, assessed by SMI (Young, 1990; Giesen-Bloo et al., 2006). The SMI is also useful as an adjunctive tool to other personality inventories when a more comprehensive psychological assessment of individuals with mental health difficulties is required.

CONCLUSION

To conclude Schema Mode Inventory (SMI) English version was translated by the researchers into Urdu language, which is the national language of Pakistan and one of the major languages in subcontinent. Our findings show that SMI Urdu version is a reliable and valid instrument which seems suitable for the assessment of schema modes. It can be used successfully both in clinical and academic settings. The SMI (Urdu) can aid in the assessment and understanding of individuals with a wide a range of psychiatric disorders and can also form the basis for Schema Focused Therapy (SFT), a therapy showing promise for a number of clinical conditions, especially Borderline Personality Disorder. However, the use of any self-report inventory inherits certain limitations, about which the clinician and researcher must be aware. This study also suggests a cross cultural study on use of SMI to further strengthen its reliability and validity and its role in the understanding of schema perpetuation and schema healing across the culture.

LIMITATIONS AND SUGGESTIONS

1. It was not feasible to subject the data to factor analysis and establish factor structure of SMI Short Form (owing to smaller number of respondents compared to number of items).
2. The participants represented non-clinical population only.
3. To determine the factor structure and psychometric properties of SMI Short Form (Urdu and English versions) for Pakistani sample, a larger representative group of respondents including non-clinical as well as clinical must be employed for factor analysis and establish construct validity of the scale.
4. Urdu translation of all the items that bear low correlations with the total score must be revised.
5. Future researchers must establish concurrent as well as discriminant validity of SMI.
6. Test-retest and equivalent form reliability must be computed for SMI on a larger sample.

REFERENCES

1. Arntz, A., Klokman, J., & Sieswerda, S. (2005). An experimental test of the schema mode model of borderline personality disorder. Journal of Behaviour Therapy and Experimental Psychiatry, 36, 226-239.
2. Bamber, M. (2008). The good, the bad and defenseless Jimmy- A single case of schema mode therapy. Journal of Clinical Psycholology and Psychotherapy, 11, 425-438.
3. Berstein, P., Arntz, A., & de Vos, E. (2007). Schema focused therapy in forensic setting theoretical model and recommendation for best clinical practice. Journal of Forensic Mental Health, 6, 169-183.
4. Giesen-Bloo, J., Van Dyck, R., Spinhoven, P., Van Tilburg, W., Dirksen, C., & Van Asselt, T. (2006). Outpatient psychotherapy for borderline personality disorder: Randomised trial of schema-focused therapy vs. transference – focused Psychotherapy. Archives of General Psychiatry, 63, 649-658.
5. Lobbestael, J., Van Vreeswijk, F., & Arntz, A. (2007). Shedding light on schema modes: A clarification of the mode concept and its current research status in Netherland. Journal of Psychology, 63, 76-85.
6. Lobbestael, J., Van Vreeswijk, M., Spinhoven, P., Schouten. E., & Arntz, A. (2010). Reliability and validity of the short Schema Mode Inventory (SMI). Behaviour Cognitive Psychotherapy, 38, 437-458.
7. Lobbestael, J., Arntz, A., & Sieswerda, S. (2005). Schema modes and childhood abuse in borderline and antisocial personality disorder. Journal of Behaviour Therapy and Experimental Psychiatry, 36, 240-253.
8. Mason, I. (1994). Techniques of translation revised: a text linguistic review of borrowing and modulation. In A. H. Albir (Ed.), Estudis 10 optimality in translation sobre la traducció (pp. 61-72). Castelló: Publicacions de la Universitat Jaume.
9. Perocelli, V., Glaser, A., Calhoun, B., & Campbell, F. (2001). Early maladaptive schemas of personality disorder subtypes. Journal of Personality Disorder, 15, 546-559.
10. Young, J., Atkinson, T., Arntz, A., Weishaar, M. (2005). The Young Atkinson Mode Inventory (YAMI-PM, 1B). New York, USA: Schema Therapy Institute.
11. Young, J. (1990). Cognitive therapy for personality disorders: A schema focused approach. Sarasto, FL: Professional Resource Exchange, Inc.
12. Young, J. E., & Klosko, J. (1994). Reinventing your life. New York, USA: Plume
13. Young, J. E. (1999). Cognitive therapy for personality disorders: A schema-focused approach. Sarasota, FL: Professional Resource Press.
14. Young, J. E., Klosko, J. S., & Weishaar, M. E. (2003). Schema therapy: A practitioner’s guide. New York, USA: Guilford Press.
15. Young, J. E., Arntz, A., Lobbestael, J., Weishaar, M. E., & VanVreeswijk, M. F., (2007). The Schema Mode Inventory. New York, USA: Schema Therapy Institute.